Diffuse large B-Cell lymphoma (DLBCL) is an aggressive form of non-Hodgkin Lymphoma (NHL) and is the most common lymphoma subtype in adults with 45,000 new cases per year. Standard therapy consists of combination chemotherapy (R-CHOP). While up to 60% of patients can be cured with chemotherapy, the remaining 40% usually see initial tumor size reduction but eventually relapse with chemoresistant disease. The reason why some patients relapse while others do not is currently not known. The mechanisms by which these tumors evolve and adapt to treatment are unknown. There are currently no biomarkers that can predict which patients will relapse. If such biomarkers could be found, patients at higher risk could be treated with more aggressive strategies and/or be monitored more aggressively with technologies to quantify minimal residual disease. We hypothesize that a systems biology approach can identify both mechanisms and biomarkers of relapse. To test this hypothesis, we will perform immunoglobulin heavy chain (IGH) VDJ-sequencing, exome-sequencing, transcriptome sequencing and DNA methylation profiling in an initial cohort of patients with clinically annotated diagnosis-relapse paired biopses of DLBCL. Using computational analysis, we will identify and validate the DLBCL relapse signature. We will validate key alterations in the relapse signature in a larger cohort of patients using targeted Sanger sequencing and targeted MassArray based methylation analysis. Using a CRISPR-Cas9 model we recently published (Kasap et al, Nature Chemical Biology, 2014), we will validate the involvement of key genes from the relapse signature in relapse-associated phenotypes such as chemoresistance. We will then identify novel biomarkers that identify patients at high risk of relapse using VDJ-sequencing, exome-sequencing, transcriptome sequencing and DNA methylation profiling in a cohort of patients that have not relapsed at least 5 years after initial diagnosis. We will generate candidate biomarkers using a computational analysis designed to identify such biomarkers and validate them using an independent validation cohort. The proposed study is in response to the RFA Biomarkers for Early Detection of Hematopoietic Malignancies (PA-12-221).